Starting Conditions of a Centrifugal Pump 1

[gilus02]

As I know two way are used for starting a Cen. pump:
Closed Delivery Valve
Open Delivery Valve

In an oil and gas plant, the starting condition is chosen by instrument&control discipline or process discipline?
The problem is we have high NPSHr (exceeding NPSHa) at the end of curve and if the pump starts with open delivery valve the pump will have NPSH problem in any starting time.
In our data sheet it is written "Automatic Start With Open Delivery Valve".
Can I change starting condition as a mechanical engineer?

Thank you all!

 

[EnergyMix]

I think it depends on the particular plant and probably, in a good organization, both departments work together.

Why can't you start with a throttled condition? Or install a startup line that runs back to the pump if there are concerns with starting with the valve wide open?

Want to know the do's and don'ts of Eng-Tips? Read FAQ731-376.
English not your native language? Looking for some help in getting your question across to others or understanding their answers? Go to forum1529.

 

[Artisi]

saying that the NPSHr at the end of curve exceeds NPSHa is fairly meaningless without a lot more information. You need to undertake a full hydraulic study of the start-up / installation / operating conditions as it is possible the pump will not reach end of curve before the system curve intersects the HQ curve.

Just because the application data sheet gives you X NPSHa and the pump data gives you Y NPSHa doesn't necessarily mean a problem.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[JJPellin]

You also need to consider the system arrangement. Most of our pumps are set up in pairs with one running and one spare. If this is the case, the second pump will auto-start on loss of flow or pressure. The standby pump will have an open discharge valve. But, it will have a check valve in the line, as well. If the pump starts up automatically, it will be running against a closed check valve until the discharge pressure of the pump exceeds the pressure of the downstream process. This is equivalent to starting up on a closed valve. As the check valve opens, it would behave like a pinched discharge valve and become fully open once normal forward flow is established. If this is your configuration, you should not have to worry about running off the right side of the curve during start-up.

On the other hand, if there is not a second pump and if there is not a check valve with full process pressure on the other side, then you will have more difficulty. If it needs to start automatically, then you won't be able to ask an operator to throttle the discharge for start-up. For most pumps, running off the right side of the curve even with inadequate NPSH does not cause significant damage. For a vertical turbine pump with many stages, it can be disastrous. You would need to provide more information about your particular pump for us to evaluate the risk properly.

For our tank transfer pumps, we would be worried about starting up with an open discharge because of the possibility of water hammer in the line. We would install a motor operated valve or control valve that would default to the closed position (or partially open position) and then open fully at a designed rate once the pump has been running for some period of time. The particular settings would be tuned to the particular application, system design, pump design, etc.


Johnny Pellin

 

[Pumpsonly]

Just to add to Johnny's comment.If you have a single pump system, the discharge valve will have to be in closed position when the pump is not running. You also have close the discharge valve before stopping the pump. Otherwise the liquid will be drained back to suction tank. Therefore you will always have to start the pump with the DV fully closed.

 

[LittleInch]

I have often found that start up conditions and issues are often not properly considered by process designers unless they have actually worked on a plant in operation and seen the issues involved with start up. As ably noted above, whether you start with open or closed valves or some sort of control valve depends on your setup and pressure conditions that exist at the pump discharge. If the pressure is very low, then there is a risk of tripping the pump due to excess current as it disappears off the curve and both excess velocity and potential for surge.

As started above, a proper study of startup and shutdown operating condition needs to be done with input and exchange of ideas and issues between all disciplines. This is often left to the HAZOP before it is really studied, but you should raise your concerns and issues with the process and controls engineers to see what needs to be done in your particular instance.

In reality pumps often survive startup which is outside their normal operating range, but good design should avoid this as sometimes the pumps don't start if you let them go off the curve on startup and then various mods need to be made by the commissioning and operations teams.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[BigInch]

Process guys try to exist in their flatline world and sometimes forget to take into account elevation changes and in fact anything outside "design condition". Problem is that you always have to get from 0 to design conditions and back without running out of the stability zone. Sometimes you will have to start with a valve closed because of too much system resistance and the high torque and discharge pressure required to get flow moving and reach design head and at other times you won't have enough resistance and will hit runout flow before system resistance builds up enough (if it ever does) to push the pump back to design flow and head. If you have those conditions you may need to start against a open, or closed block valve, or employ a control valve that opens just enough at all points along the pump's run up curve to keep it moving to design conditions within the stability margins.

Independent events are seldomly independent.